Abstracts

Rationale: Magnetic resonance (MR) volumetry studies have showed progressive contralateral hippocampal volume loss after surgery for medically intractable temporal lobe epilepsy. The relationship between the postoperative hippocampal volume loss and cognitive function has not been studied. We evaluate postoperative long-term changes of hippocampal volume (HV) correlating with cognitive functions in patients who underwent surgery for unilateral mesial temporal lobe epilepsy (MTLE) and achieved seizure freedom. Methods:  We studied 1.5-Tesla MRI before and after epilepsy surgery (19 anterior temporal lobectomy; 5 selective amygdalohippocampectomy) in 24 patients (mean±SD, 36.9±11.0years, 13-53 years) with MTLE. Serial MRI studies were performed consecutively; 1^st period, 6m-1y; 2^nd, 1-2 y; 3^rd, 2—3 y; 4^th, 3-5 y. HVs were measured according to the Wu’s method (AJNR, 2005). We compared HVs of patients with those of 14 age-matched controls. We analyzed the correlation between consecutive post-operative HVs and seizure duration, age at surgery, pre-operative both resected HVs and non-resected HVs. We defined dominant hippocampus in the hemisphere where Wada test revealed language function. We compared the postoperative changes of HVs between dominant and non-dominant hemispheres with concurrent cognitive functions. Results: The seizure duration was 26.5±11.8 years (7-52 years). The resected HV was 0.81±0.28 cm³, significantly smaller than the non-resected HV (1.32±0.37 cm³) (p<0.001). The pre-operative non-resected HVs were significantly smaller than HVs of controls. The volume reduction (%) of post-operative HVs compared to pre-operative HVs were -3.6±6.9%, -2.3±8.5, -3.6±10.2, and -5.0±9.5 at the post-operative consecutive periods. There was a significant decline of HV over the follow-up periods (p<0.01). The progressively large loss of HV at 3-5y was significantly correlated with the older age at surgery (p<0.05). Comparing dominant and non-dominant non-resected HVs, dominant HV were smaller than the non-dominant HV (p<0.05). There were consistent larger HV of non-dominant hippocampi than dominant hippocampi during postoperative periods. There was no difference of both verbal and visual memories between 10 patients with non-dominant hippocampi and 14 patients with dominant hippocampi. The verbal memory in the patients with dominant hippocampi improved after the resection of non-dominant hippocampi. At the 1-2y, there was a significant difference of improvement of verbal memory with dominant hippocampi compared to the decline of verbal memory with non-dominant hippocampi (p<0.05). Instead of the improvement of verbal memory, the dominant HV declined. Conclusions:  In patients who achieved seizure freedom after the resection of hippocampal sclerosis, the progression of hippocampal volume loss was significantly more pronounced with older age at surgery. Despite significant decline of postoperative HVs, verbal and visual memories spare as the affected level at the surgery. Earlier surgical intervention may have a lower potential risk of memory decline secondary to the postoperative hippocampal volume loss. Funding: None